For me the concept of an “Identifier”, as I have defined it formally, is at the very heart and core of what I have been looking into since late 1987.

An Identifier, in my use of the word in language, has a number of essential properties.

  • An Identifier is immutable for the life of the Information System in which it is being used.
  • An Identifier bears no attribute ¬†meaning whatsoever; it has a diametrically opposite job to do compared with an Attribute Value.
  • An Identifier bears no Sequential Order within the set of Identifiers to which it belongs; any attempt to “sort” Identifiers into some sort of Sequential Order must quite simply fail. The relative positions of the Identifiers will stay unchanged, just as Codd always insisted in respect of his Rows in his Relations.
  • An Identifier lives in a “Permanent”, in an analogous way to that in which a “Value” lives in a “Variable”, although this is as yet controversial.

But I love the obvious contrast between the idea and supporting word “Variable”, and the idea and supporting word “Permanent”. Variables embody Transience, while Permanents embody…Permanence.

As a matter of practical fact, the only way that I have found so far to represent / realise this concept is via what I call an “Identifier String”; an Identifier String, for example, can, and already has been made up of, a 9 character string of printable ASCII characters, drawn from the set of the 26 upshifted alpha characters and the 10 numeric digits; A to Z plus 0 (zero) to 9 (nine). I have resorted to using the dire Excel to help me generate a couple of thousand unique-to-set 9 character Identifier Strings, and the first one happens to be “67N8ZINA8″.

This style or format of presentation has the primary advantage that such character strings (generally) utterly FAIL to suggest to a human reader any intrinsic meaning whatsoever.

However anything at all constructed with ASCII characters is inherently riddled with Sequential Order, since each character is defined with an associated number, where such numbers precisely bear Sequential Order.

So SORTING by Identifier String simply has to be banned within any system that uses character based Identifiers. Such a system will also NOT make available Identifier Strings to human eyes; they will stay in the system’s sub-conscious or unconscious, in a form of hidden glue.

Professor Bill Roscoe, Head of the Department of Computer Science at Oxford University, and a Mighty Mean Mathematician, suggested to me that I have not being doing anything in Computer Science per se – so he did not have anybody in his Department to Supervise a DPhil that I might embark on.

He kindly suggested instead that I may have been wandering in Pure and Nominal Maths, and in Logic. ¬†I’m getting to agree with him.

That’ll do for now!